Sacral Slanting in Adolescent Idiopathic Scoliosis: Definition, Pathomechanism, and Surgical Implications: A Narrative Review

Abstract

Sacral slanting—defined as the oblique tilt of the upper sacral endplate measurable on standing anteroposterior whole-spine radiographs—is a distinctive and clinically underappreciated finding in adolescent idiopathic scoliosis (AIS). Its etiology is multifactorial and incidence among AIS patients reaches 40%. The direction and degree of sacral slanting interact with the distal lumbar curve type (L4-left vs. L4-right) to influence the pattern of postoperative radiographic changes at the distal spine and shoulder. In patients with L4-left curves and left-sided slanting, stopping fusion at L3 generally prevents coronal decompensation. Furthermore, 5-year data show that the disc wedge angle below the lowermost instrumented vertebra (LIV) does not progress when LIV is L3, and patient-reported outcomes are preserved without revision surgery. When fusion extends to L4 in patients with high sacral slanting, coronal decompensation is more frequent, yet functional outcomes remain comparable and no revision has been required in published series to date, indicating that the decompensation observed is radiographically significant but clinically tolerable within the medium term. In Lenke 1A patients with right-sided sacral slanting (L4-right type), the slanting lowers the last touching vertebra and may facilitate distal adding-on. In Lenke 2A patients with right-sided slanting, distal adding-on serves as a natural shoulder-rebalancing mechanism and may be reasonably monitored conservatively rather than reflexively treated, pending further validation in independent cohorts, unless progressive or symptomatic. Across all scenarios, the optimal distal fusion level remains an unresolved dilemma requiring further long-term prospective study. This review synthesizes the available literature to characterize sacral slanting as a promising radiographic parameter that may serve as a potentially useful adjunct in distal fusion level planning, while highlighting the need for prospective multicenter validation before its routine incorporation into surgical decision-making in AIS.

INTRODUCTION

Surgical correction of adolescent idiopathic scoliosis (AIS) warrants a careful fusion level selection to achieve optimal outcome [1,2]. For proximal level selection, factors such as proximal thoracic curve rigidity, shoulder height, and clavicle angle are routinely evaluated [3-5]. For distal level selection, the positions of the end, neutral, and stable vertebrae relative to the central sacral vertical line (CSVL) are the classical reference points [1,6]. However, a distal pelvic variable—the oblique tilt of the sacral upper endplate, termed sacral slanting—has emerged in the past decade as an additional and independent determinant of surgical outcomes at the lumbosacral junction [7-11].

Lee et al. [7] first systematically described sacral slanting in 2015 and demonstrated its relevance to postoperative coronal decompensation. Subsequent studies progressively refined its measurement, identified its associations with adjacent radiographic parameters, characterized its relationship to distal adding-on and shoulder balance across different Lenke curve types, and explored the underlying sacral morphology [12-15]. Despite this growing body of evidence, sacral slanting remains absent from most textbook discussions of AIS surgical planning. Sacral slanting may affect optimal selection of lowermost instrumented vertebra (LIV), although how this may guide level selection remains to be clarified. Moreover, while sacral slanting is a common phenomenon detected in AIS patients, it is unclear whether this is a cause or result of spinal deformity.

This narrative review synthesizes the published literature on sacral slanting in AIS. It addresses: (1) definition and radiographic measurement; (2) prevalence and directional characteristics; (3) etiology and pathomechanisms; (4) associations with curve type and adjacent radiographic parameters; and (5) surgical implications across Lenke curve types, including distal fusion level selection, distal adding-on, and shoulder balance (Table 1).

Table 1.

Summary of Key studies on sacral slanting in adolescent idiopathic scoliosis

Study	Design	No.	Curve type	Threshold	Key finding	Clinical implication
Lee et al. [7] (2015)	Retrospective case series	389	All Lenke (surgical)	≥ 3°, ≥ 4°, ≥ 5°	Prevalence 40.6%/29.6%/19.5% by threshold; 86.7% left-sided; most frequent: L4-L+left slanting (78.5%)	First systematic description. Sacral slanting is a critical consideration in LIV selection; L3 fusion safe in L4-L+left slanting.
					Decompensation 9.1% in L4 fusion vs. 0% in L3 fusion among L4-L+left slanting patients
Cho et al. [16] (2017)	Retrospective cohort	303	All Lenke (surgical)	≥ 3°, ≥ 5°	Sacral slanting correlated with pelvic obliquity (r=0.445) and lumbar curve (r=0.325)	Sacral slanting is primarily a compensatory mechanism for large lumbar curves via pelvic obliquity; LLD is not a direct driver.
					LLD not directly correlated (r=-0.034)
					8.9% showed ≥5° slanting with <3° pelvic obliquity
Joo et al. [12] (2018)	Retrospective cohort	126	Lenke 1A	≥ 3°	Prevalence 28.6%	In Lenke 1A L4-R with right sacral slanting, shorter fusion may be feasible because slanting lowers LTV; but adding-on risk requires vigilance.
					Direction concordant with L4 tilt in ~90%
					Preoperative lumbar Cobb angle independent predictor of adding-on
					In L4-R subgroup, right sacral slanting lowered LTV and shortened fusion level
Yamauchi et al. [18] (2023)	Retrospective cohort+controls	94+25	All Lenke (surgical)	-	S1 wedging angle (S1WA) significantly greater in AIS vs. controls (1.7° vs. 0.08°, p=0.002)	First CT-based evidence of S1 bony remodeling in AIS
					S1WA correlated with sacral slanting (r=0.449) and L4 tilt (r=0.347)	Structural sacral deformity contributes independently to radiographic sacral slanting
					S1WA greater in modifier C than A (2.4° vs. 0.8°, p<0.001)
Hwang et al. [13] (2024)	Retrospective cohor	96	Lenke 2A	≥ 3°	Sacral slanting did not directly predict PSI development (p=0.648 across groups)	Right sacral slanting drives distal adding-on in Lenke 2A, which acts as a natural shoulderrebalancing mechanism
					Right sacral slanting group showed largest ΔRSH improvement (-7.1 mm) and largest ΔDWA increase (+2.0°)	PSI is primarily determined by PTC size and MTC correction rate.
Yang et al. [15] (2025)	Retrospective cohort	120	Lenke 2A	≥ 3°	RSS group: DA 47.6% vs. 28.9% vs. 18.2% (p=0.02)	In Lenke 2A with RSS, stable distal adding-on associated with RSH improvement should be monitored conservatively.
					Inverse temporal relationship between increasing DWA and decreasing RSH most pronounced in RSS
					70.9% of DA stable; no revision required	Sacral slanting and postoperative RSH are risk factors for DA.
Hwang et al. [14] (2026)	Retrospective cohort 5-yr follow-up	118	Lenke 3,4,5,6 (modifier C)	≥ 5° (high), < 5° (low)	LIV L3+high slanting: DWA plateau, no progression over 5 yr	Strongest evidence to date: LIV L3 is preferred for structural lumbar curves with high sacral slanting.
					LIV L4+high slanting: C7-CSVL progressive increase (9.9→13.3 mm, p=0.035); decompensation 45.8% vs. 8.3% (p=0.024)	LIV L4 in this setting carries significantly high progressive decompensation risk
					Sacral slanting: OR 1.350 for AON, OR 1.209 for decompensation
Kanie et al. [19] (2024)	Retrospective cross-sectional study	25	Lenke 5C	-	SO>IO in both x-ray and CT measurements	SO is more appropriate than IO for assessing pelvic obliquity in Lenke 5C AIS.
					SO-CT correlated strongly with L5 tilt (r=0.884) vs. IO-x-ray (r=0.104, NS)	CT-measured SO should be used when sacral endplate is poorly visible on radiographs.
					S1 angle (4.5°)>S2 angle (1.2°)
					SO-x-ray vs. SO-CT: r=0.838

LIV, lowest instrumented vertebra; LLD, leg length discrepancy; LTV, last touched vertebra; AIS, adolescent idiopathic scoliosis; CT, computed tomography; PSI, postoperative shoulder imbalance; RSH, radiologic shoulder height; DWA, disc wedge angle; PTC, proximal thoracic curve; MTC, middle thoracic curve; DA, distal adding-on; RSS, right sacral slanting; C7-CSVL, C7 plumb line to central sacral vertical line increase; OR, odds ratio; AON, adding-on phenomenon; SO, sacral obliquity; IO, iliac obliqui.

LITERATURE SEARCH

A literature search was conducted using PubMed and Google Scholar without date restriction, using the following search terms: “adolescent idiopathic scoliosis” combined with “sacral slanting,” “pelvic obliquity,” or “wedge.” The search yielded 81 articles. After screening titles and abstracts, studies that did not directly address sacral slanting, were non-English, or were case reports, letters, or editorials were excluded, leaving 10 potentially eligible studies. Of these, one was excluded because it addressed sacral slanting exclusively in the context of hip dysplasia rather than AIS surgical planning, and one could not be retrieved in full text. A total of 8 studies were included in the final review and summarized in Table 1 (Fig. 1).

Fig. 1.

Literature search process.

DEFINITION AND RADIOGRAPHIC MEASUREMENT

Sacral slanting is defined as the angle between the horizontal line and the upper endplate of the sacrum as measured on a standing whole-spine anteroposterior radiograph (Fig. 2) [7,16]. The measurement is conceptually straightforward: the examiner identifies the superior cortical margin of the S1 vertebral body endplate and measures its inclination from the horizontal [16]. In practice, image quality is occasionally limited because the sacral upper surface can appear blurred or irregular on plain radiographs, particularly when the sacrum is anteriorly tilted relative to the x-ray beam. Lee et al. [7] reported good intraobserver reliability (intraclass correlation coefficient [ICC], 0.928) but only moderate interobserver reliability (ICC, 0.787–0.791) for this measurement, reinforcing the need for standardized measurement protocols. Cho et al. [16] subsequently confirmed good interand intraobserver agreement (ICC, 0.893 and 0.917, respectively) in a larger cohort. In addition to sagittal tilt of the sacrum relative to the x-ray beam, coronal pelvic rotation represents a further source of measurement error. Asymmetry of the obturator foramina on standing anteroposterior radiographs—a marker of pelvic rotation—can cause the sacral endplate to appear oblique independent of true sacral slanting, potentially leading to overestimation or underestimation of the slanting angle.

Fig. 2.

Radiographic measurement of sacral slanting. Sacral slanting is defined as the angle between the horizontal line (dashed line) and the upper endplate of the sacrum (straight line) as measured on a standing whole-spine anteroposterior radiograph.

No universally accepted angular threshold currently defines “significant” sacral slanting. Lee et al. [7] applied criteria of ≥3°, ≥4°, and ≥5° for sensitivity analyses, yielding frequencies of 40.6%, 29.6%, and 19.5%, respectively. Most subsequent studies have adopted ≥3° as the operational threshold, which captures approximately 40% of surgical AIS patients [12,15]. A threshold of ≥5° has been used in some studies to define “high” sacral slanting [14]. While the cut-off value of 3°–5° of sacral slanting may seem small and insignificant, Schwender and Denis [17] demonstrated that even a 3° of obliquity of sacrum can cause significant coronal decompensation of AIS curvature. At present, no single threshold can be recommended for universal clinical application. The ≥3° and ≥5° cutoffs reflect the exploratory nature of the current evidence base, each derived from different study designs and outcome definitions. Prospective studies specifically designed to identify the threshold at which sacral slanting exerts consistent and independent clinical impact are needed before standardization can be achieved.

Yamauchi et al. [18] introduced a complementary morphological parameter—the S1 wedging angle—defined as the angle between the S1 superior endplate and a line parallel to the S1–S2 intervertebral disc on 3-dimensional computed tomography (CT). S1 wedging angle measures intrinsic bony asymmetry of the first sacral segment independent of pelvic position, providing information that plain radiographs cannot [18]. S1 wedging angle was moderately correlated with sacral slanting (r=0.449, p<0.001), confirming that a component of radiographically measured sacral slanting reflects true osseous sacral deformity rather than positional tilt alone [18].

A related but distinct parameter is sacral obliquity, defined as the coronal tilt of the sacrum proper, as distinguished from iliac obliquity, which reflects the overall pelvic tilt including iliac and sacroiliac joint contributions [19]. Iliac obliquity is influenced by leg length discrepancy (LLD), iliac morphological asymmetry, and sacroiliac joint mobility, and therefore does not purely represent the sacral foundation upon which the lumbar spine rests [19]. Sacral obliquity measured on CT more accurately reflects the sacral contribution to lower lumbar tilt in Lenke 5C AIS (Table 2) [19].

Table 2.

Prevalence and directional characteristics of sacral slanting by threshold and Lenke curve type

Threshold	Overall	Lenke 1A	Lenke 2A	Lenke 3/4/5/6 (lumbar modifier C)	Direction (% left-sided)
≥ 3°	39.6%–40.6% (Lee 2015; Cho 2017) [7,16]	28.6% (Joo 2018) [12]	59.3% (Hwang 2024) [13]	-	~87% (Lee 2015; Cho 2017) [7,16]
≥ 4°	29.6% (Lee 2015) [7]	-	38.8% (Hwang 2024) [13]	-	~87% (Lee 2015) [7]
≥ 5° (“high”)	19.5% (Lee 2015) [7]	-	~20% (Hwang 2024) [13]	56.8% of structural lumbar curve cohort (Hwang 2026) [14]	~88% (Lee 2015) [7]
Predominant pattern	L4-left+left sacral slanting 78.5% of cases (Lee 2015) [7]	L4-L: 95% left L4-R: 75% right (Joo 2018) [12]	Right slanting associated with right DA (Hwang 2024) [13]	Sacral slanting direction concordant with L4 tilt in all patients (Hwang 2026) [14]	Left-sided predominance across all curve types

Dash (-) indicates data not reported for that specific subgroup in the cited study. Prevalence values represent the proportion of surgical AIS patients meeting the stated threshold. Direction refers to the side of the sacral elevation (left-sided=right ala lower, concordant with rightconvex thoracic curves in most AIS patients).

AIS, adolescent idiopathic scoliosis; DA, distal adding-on.

PREVALENCE AND DIRECTIONAL CHARACTERISTICS

Among 389 AIS patients who underwent corrective surgery, Lee et al. [7] reported a mean sacral slanting angle of 1.79°±3.23° (approximately normally distributed) and a prevalence of 40.6% using the ≥3° criterion. Cho et al. [16] replicated these findings in 303 AIS patients (mean, 1.86°±3.06°; prevalence 39.6% at ≥3°). Yamauchi et al. [18] found a mean sacral slanting of 2.6°±3.2° in a surgical AIS cohort, significantly greater than 0.3°±2.0° in nonscoliosis controls (p<0.001), confirming that sacral slanting is not simply a population-level baseline variation but a phenomenon specifically associated with AIS.

A striking directional asymmetry characterizes sacral slanting in AIS. In the studies by Lee et al. [7] and Cho et al. [16], approximately 87% of patients with sacral slanting exhibited left-sided tilt. The direction of sacral slanting is tightly coupled with the direction of L4 vertebral tilt (L4 tilt). The L4-left type (sacrum tilts left, L4 tilts left) is by far the most common pattern: in the series by Lee et al. [7], 124 of 158 patients (78.5%) with sacral slanting had a left-sided sacrum combined with a left L4 tilt. This concordant relationship suggests that sacral slanting and L4 tilt are part of the same coronal compensatory chain descending from the lumbar curve apex toward the sacropelvic junction [7].

By Lenke classification, sacral slanting was most prevalent in patients with lumbar modifier C (63.9%)—reflecting structural lumbar curves crossing the CSVL—and in Lenke types 3C, 5, and 6.7 Among Lenke 1A patients specifically, Joo et al. [12] reported a 28.6% prevalence, with directional concordance between L4 tilt and sacral slanting in approximately 90% of cases: L4-left type patients had predominantly left-sided slanting (95%), and L4-right type patients had predominantly right-sided slanting (75%). In Lenke 2A patients, Hwang et al. [13] found a 59.3% prevalence using the ≥3° criterion, the highest reported across any single Lenke subtype.

ETIOLOGY AND PATHOMECHANISM

Three nonmutually exclusive mechanisms have been proposed to explain sacral slanting in AIS: (1) compensatory pelvic response to large lumbar curves, (2) LLD, and (3) primary sacral morphological deformity.

1. Compensatory Mechanism

The most prevalent and best-supported hypothesis is that sacral slanting represents a compensatory adaptation to large lumbar curvature [16]. Cho et al. [16] demonstrated via partial correlation analysis that sacral slanting correlates significantly with both pelvic obliquity (r=0.445) and lumbar curve magnitude (r=0.325) independent of confounders (Fig. 3). The proposed chain is: large lumbar curve → pelvic obliquity → sacral tilt [16]. Yamauchi et al. [18] corroborated this by showing that S1 wedge angle (intrinsic sacral wedging on CT) was significantly greater in patients with Lenke lumbar modifier C than modifier A (2.4°±2.6° vs. 0.8°±2.0°, p<0.001), and that sacral wedging correlated with sacral slanting even after controlling for L4 tilt. This implies that chronic mechanical strain from large rotatory lumbar curves can remodel the sacral upper endplate itself, producing fixed bony asymmetry that is not purely positional [16,18].

Fig. 3.

Correlation between sacral slanting and adjacent structures. Structures that were significantly correlated in partial correlation analysis are connected with arrows.

Importantly, the compensatory relationship can also operate in the opposite direction: in a minority of patients with significant sacral slanting but minimal pelvic obliquity, the slanting may precede or cause lumbar curve development rather than result from it [16]. This bidirectionality complicates causal inference and underscores the importance of considering sacral slanting not just as an epiphenomenon but as a potential structural contributor to curve etiology in some patients [16,18].

2. Leg Length Discrepancy

LLD was hypothesized to contribute to sacral slanting via pelvic obliquity. However, partial correlation analysis by Cho et al. [16] showed no direct correlation between LLD and sacral slanting (r=-0.034, p=0.559), even in the subgroup with LLD >10 mm. LLD correlated with pelvic obliquity (r=0.123), and pelvic obliquity correlated with sacral slanting, but this indirect relationship was not sufficient to establish LLD as an independent driver of sacral slanting [16]. Therefore, routine whole-leg radiography to assess LLD appears to have limited value in explaining sacral slanting in AIS [16].

3. Primary Sacral Morphological Deformity

Yamauchi et al. [18] provided the first CT-based evidence that the S1 vertebra undergoes wedge-shaped deformity in AIS, with S1 wedge angle significantly greater in the scoliosis group than controls (1.7°±2.5° vs. 0.08°±1.5°, p=0.002). The prevalence of S1 wedge angle >5° was 16% in AIS versus 0% in controls (p=0.039) [18]. The moderate but independent correlation between S1 wedge angle and sacral slanting (after controlling for L4 tilt: partial r=0.384) suggests that sacral bone remodeling contributes to approximately one-third of the variance in radiographically measured sacral slanting [18]. This is analogous to the vertebral wedging known to occur at the apex and junctional levels of thoracic and lumbar scoliotic curves — driven by asymmetric mechanical loading during skeletal growth—now demonstrated to extend to the sacrum in patients with large lumbar curves [18].

SURGICAL IMPLICATIONS

1. Distal Adding-On

Distal adding-on is defined as progressive curve extension below the LIV, characterized by either an increase of more than 5 mm in the deviation of the first vertebra below the LIV from the CSVL or an increase of more than 5° in the angulation of the first disc below the LIV [1,12]. It is observed in approximately 2%–13% of patients after AIS correction and may necessitate revision surgery in severe cases [20,21].

In Lenke 1A AIS, Joo et al. [12] investigated adding-on in 126 patients with a minimum 2-year follow-up. Adding-on occurred in 13 patients (10.3%); the preoperative lumbar Cobb angle was the only independent predictor on multivariate logistic regression (p=0.022), while sacral slanting angle was significant on univariate analysis but did not reach independence after adjustment [12]. Subgroup analysis by L4 tilt direction revealed that in the L4-right type, right-sided sacral slanting was associated with a lower last touching vertebra (LTV) and shorter fusion level (p=0.001 for LTV level) [12]. The proposed mechanism is that right-sided sacral slanting exacerbates rightward deviation of L4, thereby lowering the LTV; when surgeons fuse to LTV–1 to preserve motion segments, the residual deforming force below the LIV is amplified, promoting adding-on [12]. In L4-left type patients, sacral slanting direction was concordant (left-sided in 95%) and the slanted sacrum stabilized the distal segment, reducing adding-on risk [12].

In patients with structural lumbar curves (Lenke types 3, 4, 5, and 6 with lumbar modifier C), the five-year follow-up study by Hwang et al. [14] provided the most granular long-term evidence to date. Among 82 patients fused to L3, multivariate logistic regression confirmed sacral slanting as an independent risk factor for the adding-on phenomenon (p=0.031) [14]. In the high sacral slanting subgroup (≥5°), the disc wedge angle (DWA) below the LIV was consistently greater than in the low slanting subgroup across all postoperative time points [14]. Critically, however, this difference did not increase further over 5 years—a plateau mechanistically explained by the 3 remaining mobile segments (L3–4, L4–5, L5–S1) collectively distributing the deforming stress imposed by the tilted sacral foundation (Fig. 4A) [14]. In the low slanting subgroup, DWA gradually decreased toward baseline over the same interval [14]. No patient in either subgroup required revision surgery, and Scoliosis Research Society 22-item questionnaire (SRS-22) scores improved comparably in both groups at 5 years (Table 3) [14]. The clinical implication is therefore that while high sacral slanting elevates the radiographic risk of adding-on when LIV is at L3, the natural course appears to be one of stabilization rather than progressive deformity, and the functional impact within the observed follow-up window was not significant [14].

Fig. 4.

Representative cases of high sacral slanting in the LIV L3 and LIV L4 cohorts. (A) A 13-year-old female patient who underwent deformity correction and fusion extending from T4 to L3. AIS curvature was classified as Lenke 3CN. LIV of L3 was selected despite lower end vertebra of lumbar curve was L4 to allow coronal compensation since sacral slanting was identified. Distal wedge angle below LIV increased after surgery as a compensatory mechanism to maintain coronal balance 1 year after surgery. Disc wedge angle below LIV demonstrated a stationary course during 5-year follow-up. (B) A 14-year-old female patient who was diagnosed with Lenke 6CN AIS underwent fusion extending from T5 to L4. L4 was selected as LIV since it was significantly tilted toward right side (26°). Her C7-CSVL was measured as 11 mm at postoperative 1 year, which exhibited an increase over time. C7-CSVL at postoperative 5 years was 26 mm. LIV, lowest instrumented vertebra; AIS, adolescent idiopathic scoliosis; C7-CSVL, C7 plumb line to central sacral vertical line increase.

Table 3.

Radiographic outcomes at 5 years by LIV level and sacral slanting magnitude [14]

Variable	LIV L3 high slanting ≥ 5°	LIV L3 low slanting < 5°	LIV L4 high slanting ≥ 5°	LIV L4 low slanting < 5°
Preoperative
Sacral slanting (°)	7.1 ± 2.2	2.8 ± 1.4	7.5 ± 2.5	3.2 ± 1.2
Lumbar Cobb angle (°)	53.6 ± 10.4	48.2 ± 9.1	58.8 ± 13.8	54.5 ± 13.9
DWA below LIV (°)	4.7 ± 3.5	5.0 ± 3.2	8.1 ± 5.9	6.7 ± 4.4
C7-CSVL (mm)	12.8 ± 8.7	15.6 ± 8.9	15.2 ± 9.4	13.2 ± 9.2
2 Years postoperative
DWA below LIV (°)	4.1 ± 2.8	3.1 ± 3.0	2.9 ± 2.7	2.4 ± 2.6
C7-CSVL (mm)	9.1 ± 7.2	7.9 ± 6.8	9.9 ± 8.6	7.3 ± 6.3
5 Years postoperative
DWA below LIV (°)	3.8 ± 2.4 ↔	2.5 ± 2.0 ↓	2.5 ± 2.9	2.9 ± 2.1
C7-CSVL (mm)	7.7 ± 7.1	8.9 ± 6.7	13.3 ± 8.4 ↑*	7.0 ± 4.8
Outcomes
Adding-on phenomenon	7.0% (3/43)	2.6% (1/39)	8.3% (2/24)	8.3% (1/12)
Coronal decompensation (C7-CSVL increase ≥ 15 mm)	23.3% (10/43)	20.5% (8/39)	45.8% (11/24)*	8.3% (1/12)
SRS-22 at 5 years	No significant difference between high and low slanting groups
Multivariate risk (sacral slanting)	OR 1.350 for adding-on (95% CI, 1.028–1.774; p = 0.031)		OR 1.209 for coronal decompensation (95% CI, 1.010–1.447; p = 0.038)

Multivariate OR: odds ratio from logistic regression per degree increase in sacral slanting for each outcome.

LIV, lowest instrumented vertebra; DWA, disc wedge angle; C7-CSVL, C7 plumb line to central sacral vertical line increase; SRS-22, Scoliosis Research Society 22-item questionnaire; ↔, plateau (no significant progression over 5 years); ↓, gradual decrease; ↑, progressive increase.

^*p<0.05 between high and low slanting subgroups within the LIV L4 group.

An important contextual nuance is that adding-on does not carry uniform clinical significance across curve types. In Lenke 2A patients with right-sided sacral slanting, Yang et al. [15] demonstrated in 120 patients followed for a mean of 52–54 months that adding-on occurred in 47.6% of the right sacral slanting group versus 28.9% and 18.2% in the neutral and left-sided groups (p=0.02) [15]. Crucially, 70.9% of adding-on cases were stable (DWA change ≤0.5° between years 1 and 2), and no patient required revision surgery. The temporal pattern—a progressive increase in DWA accompanied by a simultaneous decrease in radiologic shoulder height (RSH)—indicated that adding-on in this context functions as a natural shoulder-rebalancing mechanism rather than a complication requiring intervention [15]. These findings, first reported by Hwang et al. [13] in 2024 and confirmed by Yang et al. [15] in 2025, suggest that adding-on in Lenke 2A patients with right-sided sacral slanting should be monitored conservatively and not reflexively treated unless progressive or symptomatic.

2. Coronal Decompensation

Coronal decompensation—defined variably as a C7-CSVL increase exceeding 15 to 20 mm—is among the most consequential postoperative complications after AIS correction, with direct implications for patient-reported outcomes and the need for revision surgery [1,19,22,23]. The foundational data on sacral slanting as a determinant of coronal decompensation came from Lee et al. [7], who observed that among 124 patients with L4-left type curves and left-sided sacral slanting, coronal decompensation occurred in none of the 70 fused to L3 but in 2 of 22 (9.1%) fused to L4 at 2-year follow-up. The mechanistic explanation is that extending fusion to L4 reduces the number of distal compensatory segments to two (L4–5 and L5–S1), an insufficient buffer against the residual tilting force exerted by the slanted sacrum [7].

The 5-year longitudinal data from Hwang et al. [14] extended these findings and provided an important temporal perspective. In 36 patients with structural lumbar curves fused to L4, coronal decompensation occurred in 11 of 24 patients (45.8%) in the high sacral slanting subgroup (≥5°), compared with 1 of 12 (8.3%) in the low slanting subgroup (p=0.024) [14]. The C7-CSVL trajectory in the high slanting LIV-L4 group showed an initial postoperative increase followed by improvement in the first year, then a renewed gradual rise between 2 and 5 years—a pattern not visible in shorter-term studies (Fig. 4B) [14]. Multivariate logistic regression confirmed sacral slanting as an independent risk factor for coronal decompensation (p=0.038), while lumbar curve magnitude did not independently predict the outcome (Table 3) [14]. These findings underscore the biomechanically distinct role of the tilted sacral foundation: with only 2 remaining compensatory mobile segments below L4, the residual deforming force from the slanted sacrum is less effectively neutralized than when 3 segments remain below L3 [14].

A critical reassuring finding from the same study is that, despite the higher radiographic decompensation rate, SRS-22 scores did not differ significantly between high and low slanting groups at 5 years, and both groups demonstrated meaningful improvement across all 5 domains [14]. No patient required revision surgery during the follow-up period. This concordance between radiographic imbalance and preserved functional outcomes is likely attributable to the compensatory capacity of the young spine, the ceiling effect of the SRS-22 in this population, and the relatively modest absolute magnitude of the C7-CSVL deviation observed [14]. In other words, sacral slanting-related coronal decompensation at the five-year mark does not appear to translate into clinically significant functional deterioration or the need for reoperation within this timeframe [14]. Whether this radiographic progression will translate into clinical deterioration over a longer horizon—through accelerated adjacent segment degeneration, progressive coronal imbalance, or increased back pain into adulthood—remains an open question. Until such long-term data are available, radiographic decompensation in this setting should be interpreted as a potentially important prognostic signal rather than evidence of immediate clinical failure, and continued surveillance is warranted.

3. Postoperative Shoulder Imbalance

An unexpected but clinically important body of evidence has emerged linking sacral slanting at the base of the spine to shoulder balance at the apex. Because the lever arm from S1 to T1 spans the entire spinal column, even small sacral tilts can theoretically propagate upward through compensatory curves to influence thoracic alignment and shoulder height [13,15].

Hwang et al. [13] studied 96 Lenke 2A AIS patients stratified by sacral slanting direction (left>3°, neutral, right>3°). While sacral slanting did not directly predict postoperative shoulder imbalance (PSI; p=0.648 for PSI occurrence across groups), a striking differential emerged in shoulder height recovery over time: the right-sided sacral slanting group showed the largest improvement in RSH (ΔRSH=-7.1 mm, vs. -3.3 mm in neutral and -1.5 mm in left-sided groups; p=0.037) [13]. Concurrently, this group showed the largest progressive increase in distal wedge angle (ΔDWA=+2.0°; p=0.030) [13]. The investigators proposed that distal adding-on, driven by right-sided sacral slanting, mechanically elevates the right shoulder over time, effectively acting as a natural rebalancing mechanism [13]. The ultimate determinants of PSI development were a smaller preoperative proximal thoracic curve (p=0.019) and a higher main thoracic curve correction rate (p=0.043), confirming that sacral slanting influences the trajectory of shoulder recovery rather than the binary occurrence of PSI [13].

Yang et al. [15] confirmed and extended this hypothesis in a larger cohort of 120 Lenke 2A patients followed for a mean of 52–54 months. The right sacral slanting group consistently demonstrated the highest DWA values at all postoperative time points and the greatest reduction in RSH [15]. The inverse temporal relationship between increasing DWA and decreasing RSH in the right sacral slanting group was visually and biomechanically coherent: as the distal spine wedged rightward below the LIV, the mechanical cascade translated upward to elevate the depressed right shoulder [15]. These findings, taken together, support a paradigm shift in interpreting adding-on in Lenke 2A patients with right sacral slanting—from a complication to be prevented toward a natural adaptive mechanism that improves shoulder aesthetics and should be monitored conservatively unless progressive or symptomatic (Fig. 5) [15]. It must be acknowledged, however, that the available follow-up for this interpretation does not exceed 5 years, and the long-term consequences of persistent adding-on remain unstudied in this population. Accelerated degeneration of the discs and facet joints below the LIV, progressive mechanical loading asymmetry, and future low back pain are biologically plausible sequelae that warrant investigation in longer-term prospective studies. The designation of adding-on as “adaptive” should therefore not be equated with clinical inconsequence; rather, it argues for active, structured surveillance rather than early surgical intervention.

Fig. 5.

Mechanism of shoulder rebalancing associated with sacral slanting. (A) Preoperative imaging shows spinal malalignment with rightward tilt of the sacral base. (B) Immediate postoperative correction achieves straightening of the proximal thoracic curve, resulting in depression of the right shoulder. As the shoulders subsequently attempt to restore natural equilibrium, a laterally directed force arises at the distal junction just below the fused segments (black arrows). (C) At final follow-up, distal addingon (arrow) has developed under the persistent influence of right sacral obliquity, potentially representing a compensatory response to preserve or restore shoulder balance. The lowest instrumented vertebra level shown is representative for illustrative purposes only and does not imply a specific surgical recommendation.

While sacral slanting represents a distal determinant of shoulder balance trajectory, proximal factors—particularly UIV selection and the degree of proximal thoracic curve correction— remain the primary drivers of postoperative shoulder balance in double thoracic curve AIS. Higher UIV (T2 vs. T3) has been associated with superior proximal thoracic curve correction, reduced T1 tilt, and spontaneous improvement in shoulder balance at final follow-up, with postoperative proximal wedge angle and proximal thoracic curve/main thoracic curve ratio identified as independent predictors of PSI [3,5].

Importantly, the available evidence at this stage permits characterization of the prognostic relationships among sacral slanting, adding-on, shoulder imbalance, and coronal decompensation, but does not yet support a prescriptive surgical algorithm for Lenke 2A patients. Clinical outcomes after AIS surgery are generally favorable across nearly all reported series—reflected in preserved SRS-22 scores and the rarity of revision surgery— which may itself limit the ability of future studies to demonstrate clear between-group differences in patient-reported outcomes, even when radiographic trajectories diverge. Rather than reducing this complexity to a simple decision rule, surgeons should approach Lenke 2A patients with sacral slanting by anticipating the described radiographic patterns, counseling patients on their expected postoperative course, and maintaining structured longitudinal surveillance to detect deviation from the anticipated trajectory.

4. Dilemmas in Distal Fusion Level Selection

The evidence reviewed in the preceding subsections collectively reframes the relationship between sacral slanting and postoperative complications. Across the available studies, neither adding-on nor coronal decompensation associated with sacral slanting led to revision surgery or clinically meaningful functional deterioration within the follow-up periods examined—a finding that warrants emphasis given the frequent tendency to interpret radiographic changes as synonymous with clinical failure [7,12,14,15].

When the LIV is at L3 in patients with high sacral slanting, the DWA below the LIV is greater than in low slanting patients but does not progress over 5 years; when the LIV is at L4, radiographic coronal decompensation occurs more frequently, yet patient-reported outcomes remain comparable [14]. In Lenke 2A patients with right sacral slanting, adding-on is common but predominantly stable and functionally compensatory [13]. Taken together, these data suggest that the distal biomechanical consequences of sacral slanting—across both LIV choices and across different curve types—are largely self-limiting within the medium-term follow-up window, with the spine demonstrating an adaptive capacity that buffers against clinically severe deformity progression [7,14,15].

This does not mean that sacral slanting can be disregarded in surgical planning. Rather, awareness of sacral slanting allows surgeons to anticipate the expected radiographic trajectory and counsel patients appropriately, rather than being surprised by postoperative findings that may appear alarming but are in fact within the range of predictable adaptive responses. The key unanswered question is whether the radiographic differences that are benign at 5 years will remain so over decades: whether LIV-L4 coronal decompensation in high slanting patients evolves into symptomatic imbalance or accelerates adjacent segment degeneration in adulthood, and whether adding-on in Lenke 1A L4-right type patients stabilizes or progresses beyond the medium term [14].

The dilemma of LIV selection in the presence of significant sacral slanting therefore remains genuinely unresolved, and the current evidence base, while informative, is insufficient to prescribe a single optimal strategy across all configurations. Prospective multicenter registries with long-term follow-up, systematic documentation of sacral slanting alongside standard radiographic parameters, and patient-reported outcome measures extending into adulthood are needed to determine whether the apparent medium-term tolerance of sacral slanting-related radiographic changes is durable.

POSTOPERATIVE CHANGES IN SACRAL SLANTING

An important practical question is whether surgical correction of the lumbar curve reduces sacral slanting. Lee et al. found that mean slanted angles decreased significantly after surgery in patients with left-sided slanting (5.16° preoperatively to 3.80° postoperatively, p<0.001), consistent with the compensatory hypothesis: as the lumbar curve is corrected, pelvic obliquity decreases and the sacrum de-tilts [7]. However, the correction was incomplete in many patients: 29.9% of left-sided and 47.6% of rightsided slanting patients showed no change or a slight increase in sacral slanting angle postoperatively [7]. In 31.9% of patients, slanting decreased by ≥3° [7]. This persistent residual slanting in a significant minority may reflect irreversible S1 vertebral wedging, as identified by Yamauchi et al. [18], or underlying congenital sacral asymmetry that cannot be altered by spinal fusion alone.

The implication is that sacral slanting should be evaluated as a fixed or partially fixed baseline characteristic rather than a dynamic variable that will reliably self-correct postoperatively [7,18]. Surgical planning must therefore account for the preoperative slanting angle as a parameter that will persist—and continue to influence distal biomechanics—throughout the patient’s life [7,18].

LIMITATIONS

Several limitations of this narrative review warrant explicit acknowledgment. First, and most importantly, the evidence base specific to sacral slanting in AIS is derived predominantly from a single institution. Of the studies directly addressing sacral slanting, only 2 originated from independent research groups outside Korea; no multicenter studies on this topic have been published to date. This concentration reflects the nascent state of the field rather than selective literature inclusion, but it nonetheless limits the generalizability of the conclusions presented. Second, all included studies are retrospective in design with relatively limited sample sizes and follow-up durations not exceeding 5 years; longer-term data extending into adulthood remain unavailable. Third, one key study [14] was in-press at the time of manuscript preparation and should be interpreted as preliminary evidence pending full peer review and independent replication. Fourth, the absence of a universally validated measurement protocol for sacral slanting introduces interinstitutional variability that may limit reproducibility across centers. Taken together, these limitations underscore that sacral slanting, while a promising radiographic parameter, has not yet achieved the level of evidence required for incorporation into standardized surgical planning algorithms, and that prospective multicenter validation remains an essential next step.

CONCLUSION

Sacral slanting is a prevalent, clinically meaningful radiographic finding in surgical AIS patients that deserves routine assessment alongside classical parameters such as end vertebra, neutral vertebra, and stable vertebra [7,14,16]. Its measurement on standing AP whole-spine radiographs is simple and reproducible. Its directional characteristics—predominantly left-sided, concordant with L4 tilt, and most common in curves with lumbar modifier C—reflect both compensatory spinal mechanics and intrinsic S1 vertebral remodeling [7,16,18].

Surgically, the current evidence supports 3 broad conclusions. First, in patients with structural lumbar curves (Lenke types 3, 4, 5, and 6 with lumbar modifier C), neither L3 nor L4 fusion led to revision surgery or clinically meaningful functional deterioration within medium-term follow-up despite their differing radiographic trajectories: L3 fusion produces a nonprogressive DWA plateau even in patients with high sacral slanting, whereas L4 fusion carries a substantially higher risk of radiographic coronal decompensation. Longer-term surveillance is warranted, as the functional tolerance of these radiographic differences into adulthood remains unestablished [7,14]. Second, in Lenke 1A patients with L4-right type curves, right-sided sacral slanting lowers the LTV and amplifies the residual deforming force when fusion stops at LTV-1, increasing adding-on risk—a mechanism distinct from that operating in structural lumbar curves and not to be conflated with it [12]. Third, in Lenke 2A patients with right-sided sacral slanting, adding-on is common, predominantly stable, and associated with progressive shoulder rebalancing, and should be interpreted as an adaptive rather than a pathological phenomenon unless symptomatic or progressive beyond the stable phase [13,15]. Across all scenarios, the optimal distal fusion level in the presence of significant sacral slanting remains a genuinely unresolved dilemma that requires further prospective, multicenter, long-term investigation.

Future prospective multicenter studies are needed to address several unresolved questions. First, long-term patient-reported outcomes extending into adulthood are essential to determine whether the radiographic differences observed at 5 years—including coronal decompensation in LIV L4 patients and persistent adding-on in Lenke 2A patients with right sacral slanting —translate into clinical deterioration over time. Second, the natural history of adjacent segment degeneration below the LIV in the presence of high sacral slanting requires dedicated investigation, as accelerated disc and facet degeneration remain plausible but unstudied sequelae. Third, prospective validation of sacral slanting thresholds is needed to establish which degree of obliquity carries consistent and independent clinical significance. Addressing these priorities will determine whether sacral slanting can be meaningfully integrated into formalized AIS surgical planning algorithms to refine fusion level selection and improve long-term outcomes.